A Galaxy of New Worlds
WASHINGTON, D.C.—The American Astronomical Society meeting, held here this week, was officially the largest congregation of astronomers (3,400 of them) in history—the most extraordinary collection of cosmic knowledge that has ever gathered together with the possible exception of when Isaac Newton dined alone. The breadth of topics was astounding, but the one that stood out was the study of planets beyond our solar system. The Kepler space observatory's discoveries got the most attention, but I filled an entire notebook with other exoplanetary findings and theorizing.
What makes exoplanets so much fun is not only that they bring astronomers closer to their dream of finding another Earth out there, but that the planets never fail to surprise. Every new planet-finding technique brings new looks of astonishment. For the Doppler-shift technique that discovered the first exoplanet around a sunlike star, the surprise was that planets can orbit so close to their stars. For the transit technique now being employed by Kepler, it's that the planets have unexpected densities.
Jupiter- and Saturn-mass planets in tight orbits are less dense than expected—as low as about 0.1 gram per cubic centimeter. Styrofoam was the metaphor of choice at the meeting, perhaps because sleep-deprived astronomers and journalists had coffee on their minds. One reason for such fluffy planets might be that stars heat them up and puff them out, but theorist Dimitar Sasselov of Harvard University says he's doubtful. For one thing, stellar radiation warms the top of the atmosphere, whereas to puff up a planet you need to inject the energy deep inside. Tidal heating—whereby the star's gravitational forces knead the planet and warm it up as surely as bending a paper clip back and forth makes it hot—could do the trick for planets that have noncircular orbits, but most don't. Another idea is that the planet may retain more of the heat generated during its formation.
For more modest worlds like Uranus and Neptune, the problem is that the planet mass and size don't form a continuum. There's a distinct gap between these planets and the larger ones. One idea is that those smaller worlds used to be bigger and their outer layers of gas boiled off after billions of years of relentless stellar heating. Brian Jackson of NASA's Goddard Space Flight Center calculates that a Saturn-like planet close to its star could shed 100 Earth-masses over a few billion years.
At the very lowest end of known planetary size, approaching Earth in size, the density mystery is the wide range of density values, which hints at a diversity of composition scarcely imagined a decade ago.
Properties such as density reflect how planets form and evolve, but reconstructing this history has been hindered by observational selection effects. In perhaps my favorite talk at the meeting, Scott Gaudi of Ohio State University described how neither the Doppler nor the transit technique so far has been able to take a representative sample of planets—partly because they are inherently sensitive to large, close-in planets, partly because astronomers have organized their observations to maximize their chances of discovering planets rather than getting good statistics. So astronomers don't know whether they're seeing the rule or the exceptions.
Gaudi's own technique of gravitational microlensing—whereby a planet reveals itself by passing in front of a background star and temporarily magnifying the starlight—avoids the second problem (sample selection). It monitors millions of background stars and imposes no prejudice about what might pass in front of them. It doesn't avoid the first problem (biased sensitivity), but complements other techniques by picking up planets on more distant orbits. That's where giant planets, even those that are on tight orbits, are thought to originate.
Gaudi's µFUN survey—a shoestring operation that combines the efforts of professional and amateur astronomers—has found five planetary systems so far, one with a pair of planets analogous to Jupiter and Saturn. If each of the 13 stars in the survey had planets, a total of 18 should have turned up, taking into account the biases of the technique.
Venturing to do statistics with this tiny sample, Gaudi and his colleagues conclude that one in three stars has a planet—whereas the Doppler technique implies only one in 20. If so, most planets stay in the distant orbits where they form rather than migrate inwards. Moreover, the occurrence of one planetary pair out of six cases hints that such pairs occur in one of six systems—which makes our solar system a minority, but not a complete outlier.
As theorists struggle to interpret the data, observers continue to improve their instrumentation. Ron Walsworth of the Harvard-Smithsonian Center for Astrophysics described a new technology that could allow the Doppler technique, still the gold standard, to detect Earths. The technique looks for a planet's slight tug on its host star; Earth causes the sun to move about 10 centimeters per second. Today's state of the art can detect a velocity of perhaps 1 meter per second. One problem is the lack of a sufficiently stable and precise yardstick for measuring spectral lines and their minute Doppler shifts. Walsworth argued that the answer may be laser frequency combs, which sounds like a scary cousin of laser hair removal, but is in fact an ultraprecise frequency standard created by rapidly pulsing lasers.
Despite these advances, Kepler is still the best hope for finding the first true Earth analogue. A planet in an Earth-like orbit only goes in front of its star once a year, by definition, so it will take Kepler a few more years to spot one. In the meantime, it will generate plenty of excitement—and not just about planets. By monitoring millions of stars for planets, the observatory can't help but glean information about those stars. Natalie Batalha of San Jose State University used Kepler to tackle an old question: how typical is the sun? The sun's brightness oscillates by a fraction of a percent. Are other stars of its spectral type similarly serene?
The answer appears to be yes. The 20 brightest G-type stars are just as stable. Indeed, of 43,000 G-type stars that Batalha examined, two-thirds are, if anything, less prone to activity such as solar flares. Their relative tranquility, as much as the discovery of planets, improves the prospects for life elsewhere in the Milky Way.